We designed two experiments to investigate subsyllabic unit preference in reading Pinyin in Chinese kindergarteners. Pinyin is an alphabetic transcription used in Mainland China to assist children in learning to read Chinese characters. In Pinyin instruction, teachers rely heavily on onset-rime division. Spoken Chinese, however, encourages body-coda division. In Experiment 1, a Pinyin analogy task was employed in which children were asked to read 16 groups of CVC pseudo-Pinyin syllables with different shared subsyllabic units (i.e., body, rime, and onset-coda). In Experiment 2, a short-term recall task was administered using 16 lists of three CVC pseudo-Pinyin syllables and recall error patterns were analyzed. In both experiments, a clear preference for the rime unit in comparison to the body unit was found. More accurate performance was observed in both Experiments 1 and 2 when tone was matched across the stimuli. Children retained more rime than body units regardless of onset sonority type in Experiment 2. These results were discussed in terms of the strong influence of the instructional approach on young children's subsyllabic unit reference in decoding written syllables.